
	Where results make sense

Topic: Feature vectors

Related Topics

vector spatial

vector field

vector graphics

vector calculus

normed vector space

vector bundle

Four vectors

vector biology

topological vector space

unit vector

vector processor

In the News (Mon 21 Dec 09)

Northern Trust

Marvin Harrison

Peter Chernin

Gary Locke

Match Play

Penelope Cruz

Mickey Rourke

AR Rahman

Danny Boyle

Hugh Jackman

	United States Patent Application: 0030204492
		The feature vector is projected to a low dimension document feature vector, and the documents are indexed according to the low dimension document feature vectors.
		Each feature vector is projected to a low dimension document feature vector, and the documents are indexed in a document index according to the low dimension document feature vectors.
		This certainty vector is an analog of the vector space 102 representation of documents 101, and is then subjected to the same projection (SVD, LSA etc.) applied to the document feature vectors 102 to produce the low dimension query certainty vector 107.
	appft1.uspto.gov /netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PG01&p=1&u=/netahtml/PTO/srchnum.html&r=1&f=G&l=50&s1="20030204492".PGNR.&OS=DN/20030204492&RS=DN/20030204492 (3893 words)

	Model Revision
		This section examines one model revision algorithm [5] to illustrate the process of revising the feature space using feedback from the user and the ground truth results.
		The corresponding feature vector is extracted from the example, and the K best matches are retrieved using a Euclidean metric.
		The K results whose feature vectors are closest to the target feature vectors are then returned to the user for visual inspection or further processing.
	www.research.ibm.com /networked_data_systems/9auc/model/node18.html (412 words)

	[No title]
		Given any new feature vector from the same population, not necessarily in the sample, it can be recognized as belonging to one of the classes based on its similarity to the prototypes.
		Another useful example is a feature vector taken from patients that records various activities and characteristics of the patient.
		Learning from labeled feature vectors is called supervised learning because the classes are already known and the PR system adjusts its parameters so that its output is the correct class label (or codeword) for each input (labeled) feature vector in the sample.
	www.cs.unr.edu /~looney/cs479/cs4791.htm (1997 words)

	SONBRG (Site not responding. Last check: 2007-10-16)
		Once vectors for all the words in the corpus are calculated, the dimension of the space is increased through random selection of an additional set of feature words.
		These vectors represent the relationship between the two sets of feature words, so extending the initial seed vectors to eight dimensions is simply a matter of filling in the remaining components with values representing the word relationships already calculated, as shown in matrix BT.
		This process of incrementing the dimension of the space and recalculating word vectors continues until the dimensionality is sufficient to encode relationships amongst the total number of words in the corpus.
	www.rl.af.mil /div/IFE/PPapers/SONBRG_PAPER/sonbrg.html (802 words)

	[No title]
		Feature vectors that are farther away from the cluster center should not have as much weight as those that are close.
		2.1 The k-means Algorithm The k-means algorithm assigns feature vectors to clusters by the minimum distance assignment principle [5], which assigns a new feature vector x(q) to the cluster c(k) such that the distance from x(q) to the center of c(k) is the minimum over all K clusters.
		When a feature vector is of equal distance from two cluster centers, it weights the same on the two clusters no matter what is the distribution of the clusters.
	www.cse.unr.edu /~lzhang/fuzzyCluster/paper/fuzzypaperMayNoK.doc (4027 words)

	World Intellectual Property Organization (Site not responding. Last check: 2007-10-16)
		Prior to retrieving another group of acoustic feature vectors, similarity measures are computed for the first group of acoustic feature vectors in relation to each of the acoustic models employed by the speech recognition system.
		It is envisioned that the number of acoustic feature vectors associated with the first group and the number of acoustic models loaded into the memory workspace should be selected to optimize use of the available memory space.
		It is envisioned that the first group of acoustic feature vectors is removed from the memory workspace prior to loading the subsequent group of acoustic feature vectors into the memory workspace.
	www.wipo.int /ipdl/IPDL-CIMAGES/view/pct/getbykey5?KEY=03/90203.031030&ELEMENT_SET=DECL (8600 words)

	[No title]
		As was discussed previously, feature vectors that are farther away from the prototypical vector for a cluster should not have as much weight as those that are centrally and densely located.
		To give deviant feature vectors the same weight in the averaging of a cluster as the more centrally and densely situation ones causes the prototype obtained by weighted averaging to no be as representative of its cluster as is should be.
		Rather than let a feature vector belong to a single cluster with a truth value of 0 (false, which means it does not belong) or 1 (true, it does belong), the fuzzy weights permit an in-between truth value of belonging.
	www.cs.unr.edu /~looney/cs479/cs4795.htm (1701 words)

	Binary Image Analysis: Character Recognition
		Using these moments, the invariant nature of the feature vectors can be taken a step further by applying a set of moment functions called, "Hu functions" which result in translation, rotation and scale invariant features.
		Once the feature vectors are calculated for the training set, a set of covariance matrices can be generated that characterize the feature vector data for each sample character.
		Another critical implementation feature is the ability to split characters that were merged as a consequence of the connected component analysis.
	scv.bu.edu /GC/ickey/p1/p1.html (763 words)

	Preliminaries
		Feature: The next higher abstraction level for representing images is at the feature level.
		An image feature is a distinguishing primitive characteristics or attribute of an image [8].
		These feature vectors can be predefined and pre-extracted, or user-defined and pre-extracted, or even user-defined and extracted at query time.
	www.research.ibm.com /networked_data_systems/spire/whitepaper/node2.html (699 words)

	Chapter 3 Preprocessing Of The Speech Data
		All pattern vectors are warped against a reference pattern vector of the same category which has the same number of feature vectors as there are frames in the input layer of the neural network.
		Subsequent feature vectors for the warped pattern vector are chosen such that the nth feature vector is that feature vector from the input pattern vector B closest to the nth feature vector of the reference pattern vector.
		Successive feature vectors of the time aligned pattern vector are then chosen such that the Euclidian distance between each of them is as close to L as possible.
	www.moonstar.com /~morticia/thesis/chapter3.html (2149 words)

	Publications List
		A central feature of this approach is that the effective dimensionality of the latent space (equivalent to the number of retained principal components) is determined automatically as part of the Bayesian inference procedure.
		Magnification factors specify the extent to which the area of a small patch of the latent (or `feature') space of a topographic mapping is magnified on projection to the data space, and are of considerable interest in both neuro-biological and data analysis contexts.
		The salient features of coupled transport have been assessed and demonstrated to be fully consistent with experimental data: it has been shown that transport matrices with relatively large off-diagonal components can lead to small apparent perturbations of the density, when the energy balance is perturbed, whilst still affecting the thermal transport considerably.
	www.research.microsoft.com /~cmbishop/publications_abs.htm (13203 words)

	[No title] (Site not responding. Last check: 2007-10-16)
		Graph spectral feature vectors are calculatedfirom the leading eigenvalues and eigenvectors of the unweighted graph adjacency matrix.
		The length of the vectors are determined by the number of leading eigenvalues, and the order of the components of the vectors is the order of the eigenvectors.
		From the first row to the fifth row, we show the experimental results obtained when the spectral feature vectors are constructed from ordered eigenvalues, the cluster volumes, the cluster perimeters, the cluster Cheeger constants, the shared perimeters and the cluster distances.
	www-users.cs.york.ac.uk /~luo/paper1.txt (2419 words)

	SMR Abstract (Site not responding. Last check: 2007-10-16)
		Purpose: Volume calculations by vector decomposition suffer when there is poor separation of the prototype vectors in feature space and when tissues not represented by prototype vectors are present in the image.
		The centroid vector and covariance matrix of the distribution derived from a hand-drawn pure tissue mask was used to describe each tissue prototype.
		Comparing 2D projections to density-vs.-T2 feature spaces, grey and white matter volumes were affected by the amount of CSF (i.e., dependent on the source images for the binary mask) much more in the latter case than in the former.
	www.cmrr.umn.edu /~ebaker/smr_abs.html (378 words)

	Intonation-Based Speaker Recognition (Site not responding. Last check: 2007-10-16)
		The two vectors belonging to this cluster spanned by the greatest distance are then selected, and the cluster is split into two new clusters by reassigning every vector in the cluster to one of the two aforementioned vectors.
		That is, for each feature vector associated with the speaker, the nearest centroid in the speaker's codebook is found, and the scaled mahalanobis distance to that centroid is calculated.
		These vectors were split into testing sets of 210 vectors/reader and training sets comprising the remainder of the vectors (training data sets varied from reader to reader between 276 and 664 training vectors).
	www.cs.princeton.edu /~mdhoffma/prosody/prosody.htm (1304 words)

	Sean Landis' Fall 718 Context-Based Image Retrieval Project Page
		For each image, a feature vector is generated such that each element of the vector represents the percentage of a color quantum found in the image.
		A primitive feature is a low-level or statistical attribute of an image such as an object boundry or color histogram.
		Since in content based visual databases, all items (images or objects) are represented by pre-computed visual features, the key attribute for each image will be a feature vector which corresponds to a point in a multi-dimensional feature space; and search will be based on similarities between the feature vectors.
	www.nbb.cornell.edu /neurobio/land/OldStudentProjects/cs718/fall1995/Landis (6695 words)

	NLM/CEB - A Prototype Content-based Image Retrieval System for Spine X-rays
		At retrieval time, a feature vector q is derived from the user's query, and the database of feature vectors is navigated to locate feature vectors similar to q.
		An outstanding problem in the extraction of feature vectors from the raw boundary data is, achieving a significant data reduction while simultaneously preserving the shape characteristics essential for the end use of the database.
		If this shape is not adequately preserved in the feature vector calculation, the extent, or even the presence, of the osteophyte may not be detectable for retrieval.
	archive.nlm.nih.gov /pubs/long/cbms2003/cbms2003.php (3121 words)

	SMOTE-N
		A matrix defining the distance between corresponding feature values for all feature vectors is created.
		This equation is used to compute the matrix of value differences for each nominal feature in the given set of feature vectors.
		To generate new minority class feature vectors, we can create new set feature values by taking the majority vote of the feature vector in consideration and its k nearest neighbors.
	www.cs.cmu.edu /afs/cs/project/jair/pub/volume16/chawla02a-html/node16.html (398 words)

	Gallery (Site not responding. Last check: 2007-10-16)
		The horizontal axes index the weight vector coefficients of the 546 training examples; the vertical axes show their values.
		The weight vector coefficients were uniformly initialized to a value of one.
		Sparse feature vectors lead to faster convergence of the multiplicative updates for classification by mixture models.
	www.cis.upenn.edu /~lsaul/gallery.html (800 words)

	Normalization of speech by adaptive labelling (US4926488)
		The string of normalized vectors or the string of associated prototypes (or respective label identifiers thereof) or both provide output from the acoustic processor.
		means for generating a first modified feature vector signal having a modified feature value, said modified feature value being related, by a modification function, to the feature value of a first feature vector signal in the series of feature vector signals;
		means for outputting the identification value of the prototype vector signal associated with the second modified feature vector as a coded representation of the second feature vector signal.
	www.delphion.com /details?pn=US04926488__ (642 words)

	School of Computing Sciences (CMP) (Site not responding. Last check: 2007-10-16)
		The aim of this project is to further reduce the bit-rate needed for encoding speech feature vectors whilst retaining good recognition performance on both clean and noisy speech.
		The approach to compression in this project is to exploit the temporal correlation which exists in the feature vector stream to reduce the amount of data needed for representation.
		The current algorithm receives a block of static feature vectors (typically 8 x 14 dimensional MFCC vectors), groups them together and applies a discrete cosine transform (or similar) across the temporal axis of the block.
	www.sys.uea.ac.uk /web/research/showproject.jsp?labid=2&groupid=6&projectid=30 (294 words)

	Experimental Results
		All four possible vectors are used in the training set, which is why perhaps this is a simplistic example.
		Some current research deals with using AI techniques to determine the components of the feature vectors, and I think this would be an interesting area of study.
		One problem is that I can easily envision feature vectors with attributes on which no linear order is defined; these components would have to be excluded from the attribute vector calculations.
	sean-janus.optionpc.com /truman/summer-research/results.html (749 words)

	Selection of Features (Site not responding. Last check: 2007-10-16)
		The theory is that two frames with very similar feature vectors would represent the same phoneme.
		Conversely, two feature vectors from different phonemes would not be similar.
		At any rate, it is fairly straightforward to build long lists of these feature vectors, given utterances (wave files) and their phonetic transcriptions (phonetic lola files).
	www.cse.ogi.edu /~ldcolton/vitae/cse553/manual/node74.html (293 words)

	L. Schomaker c.s., Abstracts
		Although the use of feature extraction and abstract coding can alleviate the problem to some extent, the range of idiosyncratic handwriting properties is so wide that an early-stage narrowing of the domain of possible recognition solutions is necessary.
		Therefore, those stroke features which show the mot invariant pattern are probably related to the parameters of the higher-level representation, whereas the more noisy features are probably related to the parameters derived at the lower levels (topdown hierarchy).
		Based on a set of handwritten words, three features are determined: a {\em cursivity index} $c$, which indicates the tendency of a writer to write cursive, and two distance measures $d_c$ and $d_h$.
	www.ai.rug.nl /~lambert/publications/abstracts.html (8152 words)

	Method for high-dimensionality indexing in a multi-media database (US5647058)
		The feature vector is then transformed in a manner such that the similarity measure is preserved and that the information of the feature vector +E,rar v+EE is concentrated in only a few coefficients.
		The entries of the feature vectors are truncated such that the entries which contribute little on the average to the information of the transformed vectors are removed.
		An index based on the truncated feature vectors is subsequently built using a point access method (PAM).
	www.delphion.com /details?pn=US05647058__ (602 words)

	Sentiment classification on customer feedback data: noisy data, large feature vectors, and the role of linguistic ...
		Sentiment classification on customer feedback data: noisy data, large feature vectors, and the role of linguistic analysis.
		We show that by using large feature vectors in combination with feature reduction, we can train linear support vector machines that achieve high classification accuracy on data that present classification challenges even for a human annotator.
		Michael Gamon, 2004: Sentiment classification on customer feedback data: noisy data, large feature vectors, and the role of linguistic analysis.
	research.microsoft.com /research/pubs/view.aspx?type=Publication&id=1138 (139 words)

	Machine Learning Methods (Site not responding. Last check: 2007-10-16)
		When predicting the class of a vector, a decision tree passes a vector down the tree from the root to a leaf.
		At each node, the decision tree examines one feature of the vector to determine which branch the vector should recursively travel down.
		When a subset's elements all belong to the same class or the amount of information in the subset is statistically insignificant, a leaf is formed, whose classification is equal to the majority classification of the subset.
	www.cs.washington.edu /homes/djp3/Compbio/PSB2002/paper/node6.html (301 words)

	Simlifying the Feature Vectors (Site not responding. Last check: 2007-10-16)
		We now have feature values for a sequence of frames, each representing a time slice.
		So we can't go straight from a single feature vector to a phoneme..
		C256), by measuring the similarity of a feature vector with templates in the codebook.
	www.cee.hw.ac.uk /~diana/nl/l16sh/node11.html (95 words)

	ICSLP-2000 Abstract: Ramabhadran et al. (Site not responding. Last check: 2007-10-16)
		Selection of acoustic features for robust speech recognition has been the subject of research for several years.
		In the past, algorithms that use feature vectors from multiple frequency bands [1], or employ techniques to switch between multiple feature streams [2] have been reported in the literature to handle robustness under different acoustic conditions.
		In this paper, we propose a likelihood-based scheme to combine the acoustic feature vectors from multiple signal processing schemes within the decoding framework, in order to extract maximum benefit from these different acoustic feature vectors and models.
	www.isca-speech.org /archive/icslp_2000/i00_3913.html (308 words)

Try your search on: Qwika (all wikis)